Ketones are useful organic solvents for painting, agricultural, and other industries. Preparation of ketones from carboxylic acids with metal oxide catalysts at high temperatures has long been known. Catalysts suggested for this process include oxides of thorium, zirconium, titanium, calcium, barium, cerium, chromium, aluminum, lanthanum, neodymium, samarium, etc. However, with all of these catalysts, there is a problem with obtaining selective formation of a mixed ketone derived from a mixture of carboxylic acids. Due to the nature of the ketonization reaction, each molecule of ketone is produced from two molecules of carboxylic acids. Consequently, a mixture of two carboxylic acids gives three possible ketone products. Two symmetrical ketones are derived from the same type of carboxylic acid, and the third one, a mixed ketone, is derived from two different carboxylic acids. When the mixed ketone is the desired product, there is a need to increase the selectivity to the mixed ketone, i.e., to produce a smaller amount of the other two symmetrical ketones.
While a number of catalysts have been suggested in the literature for the ketonization, there is little teaching in the art on how to improve the selectivity to the mixed ketone. One such example is given in U.S. Pat. No. 4,950,763. However, the selectivity to the mixed ketone in the '763 patent is close to the statistically expected value. In the current invention, we have discovered certain catalysts that can produce mixed ketones with a selectivity much higher than the statistically expected value.
One factor determining selectivity to the mixed ketone is the molar ratio of the two acids used for its preparation. The theoretical selectivity is based on the assumption that two components, A and B, have equal reactivity and equal probability of forming either a symmetrical or a mixed product. If A and B are present in equal amounts, the probability of forming product AA, AB, and BB is 25%, 50%, and 25%, respectively. In general, the probability of forming a mixed product AB from a mixture of A and B can be calculated by trivial formulas, R/(R+1) based on starting material B, and 1/(R+1) based on starting material A, where R is the molar ratio of A:B. In the above example, since the amount of A and B is equal, one half of A goes to AA, and another half goes to AB. Thus, the theoretical selectivity to AA and AB is 50% each, based on A. Similarly, BB and AB are formed in 50% selectivity based on starting material B.
The '763 patent reported 99% selectivity to the mixed ketone product, such as propiophenone. However, this high selectivity is due to the general difficulty in forming one of the symmetrical products, benzophenone, regardless of the catalyst used. Benzophenone is difficult to make from benzoic acid in the described process, so the only expected product with the phenyl group would be propiophenone.
In the propiophenone example of the '763 patent, the molar ratio of butyric to benzoic acid is 5.88:1. The selectivity to the mixed ketone, propiophenone, is high, 99%, based on benzoic acid, which is above the statistically expected value of 85.5%. As mentioned above, this fact is due to the natural difficulties in making the benzophenone molecule, regardless of the type of catalyst used. However, from the '763 patent, it is not clear what the observed selectivity to propiophenone was, based on the other acid, butyric acid. In any case, the high ratio of butyric to benzoic acid used in this example dictates the theoretically expected selectivity to be 14.5%, which is very low. The reason for spending so much butyric acid is that butyric acid makes dipropyl ketone more readily than the mixed one, propiophenone. Ideally, it would be desirable to have a catalyst that makes only the mixed ketone from a pair of acids, using a 1:1 molar ratio, or at least provides selectivity to the mixed ketone higher than expected statistically on the basis of either acid.
In another example, but one where a symmetrical product can be formed, the '763 patent reported 72% selectivity to the mixed ketone, methyl isopropyl ketone, made from isobutyric acid and acetic acid, on an isobutyric acid basis, at a 1.5:1 molar ratio of acetic acid to isobutyric acid. This selectivity is much lower than the 99% discussed above, but it is better than the theoretical selectivity of 60%. However, the selectivity to the mixed ketone on an acetic acid basis is not reported. Since acetic acid is used in excess to isobutyric acid, it is expected that the symmetrical ketone, in this case acetone, is made in a larger amount than the desired product, methyl isopropyl ketone.
One of the purposes of the current invention, therefore, is to find catalysts that provide high selectivity to a mixed ketone from a mixture of carboxylic acids and that make less of the symmetrical ketones.